Projects and Prototypes
Lightweight tools for adaptive learning and human-centred AI.
I build lightweight, AI and data-driven tools for human-centred AI, adaptive learning, physiological machine learning, and applied ML systems. My project work focuses on practical prototypes that are easy to use, transparent, and low-cost to maintain.
Project Ideas
📝
Cognitive Load Check
A simple web-based self-assessment tool for tracking perceived mental effort, task difficulty, confidence, and fatigue across study or work sessions.
Output: A personal dashboard showing cognitive load patterns over time.
🎯
Adaptive Study Planner
A rule-based study planner that adjusts review strategy based on quiz performance, confidence, perceived difficulty, and recent workload.
Output: Suggested next action, such as review, retrieval practice, rest, or increased difficulty.
⌚
Physiological Learning Journal
A personal learning journal that lets users manually input or import simple physiological indicators such as sleep duration, resting heart rate, HRV, and subjective stress.
Output: Visual correlation between learning performance, perceived load, and well-being indicators.
🤖
AI Workflow Load Evaluator
A lightweight framework for evaluating whether an AI tool actually reduces user cognitive load in a workflow.
Output: A rubric and dashboard for comparing workflow conditions before and after AI assistance.
Project Principles
⚡ Lightweight before complex
🧭 Rule-based before heavy training
🔒 Privacy-aware by default
🧑🎓 Useful for learners and professionals
🧪 Research-oriented, but practical
📤 Exportable and reproducible
Tools and Methods
My current project work emphasises simple and maintainable tools such as Python, Streamlit, static web apps, JavaScript, CSV-based data workflows, public datasets, and reproducible notebooks.
Arkaan Izhraqi 